中国科学院数学与系统科学研究院

科研进展与学术交流报告会

（第46期）

报告人一： 潘文亮 副研究员（系统科学研究所）

题目一：Region-based Functional genome-wide association detection for imaging response

摘 要：Advancements in data acquisition technology have fueled the growth of brain imaging genetic studies, which seek to uncover connections between brain images and genetic markers. Nevertheless, persistent challenges such as misalignment, region heterogeneity, and registration errors necessitate innovative solutions. In this paper, we introduce the region-based functional genome-wide association detection (rfGWAS) method to address these issues. By focusing on small regions rather than individual voxels, rfGWAS streamlines computation while preserving the detection of meaningful associations. Theoretical analysis confirms that rfGWAS adheres to the independence-zero equivalence principle and reliably identifies significant region sets. Moreover, its test statistic effectively controls Type-I errors under null hypotheses and attains a probability of 1 for rejecting alternative hypotheses. Simulation results underscore the method's effectiveness, and its application to hippocampus surface data from the ADNI study demonstrates its potential. rfGWAS emerges as a promising solution for uncovering region-based associations in brain imaging studies, mitigating critical shortcomings of existing approaches.

报告人二： 盛赢 助理研究员（应用数学研究所）

题目二：Sequential Data Integration Under Dataset Shift

摘要：With the rapidly increasing availability of large-scale and high-velocity streaming data, efficient algorithms that can process data in batches without requiring expensive storage and computation resources have drawn considerable attention. An emerging challenge in developing efficient batch processing techniques is dataset shift, where the joint distribution of the collected data varies across batches. If not recognized and addressed properly, dataset shift often leads to erroneous statistical inferences when integrating data from different batches. In this paper, two shift-adjusted estimation procedures are developed for updated estimation of the parameter in the presence of dataset shift. Under prior probability shift, we can obtain parameter estimation and assess the degree of dataset shift simultaneously.

时  间：2023.11.3(星期五), 10:40-12:00

地  点：南楼204会议室/腾讯会议374-6743-0937